Central banks analyze a wide range of data to obtain better measures of underlying inflationary pressures. Factor models have widely been used to formalize this procedure. Using a dynamic factor model this paper develops a measure of underlying inflation (UIG) at time horizons of relevance for monetary policymakers for both CPI and PCE. The UIG uses a broad data set allowing for high-frequency updates on underlying inflation. The paper complements the existing literature on U.S. 'core' measures by illustrating how UIG is used and interpreted in real time since late 2005.